/ 2 min read

AI in Discovery: Your Prompts May Show Intent

A new decision from the Southern District of Florida involves the intersection of discovery and AI. In Fontainebleau Fla. Hotel, LLC v. Jacob, the plaintiff-hotel is suing an individual defendant in a dispute over the defendant’s public statements about the hotel. The court had already struck the defendant’s pleadings. The court reopened discovery to determine how much money the hotel should recover, including punitive damages.

The hotel asked for the defendant’s AI chats. Its argument was simple. If the defendant used AI to write or polish statements about the hotel, its employees, and its lawyers, the prompts could show what the defendant was trying to accomplish, and the finished statements might not. 

The court agreed that those chats were fair game, because intent matters for punitive damages under Florida law. The court refused to let the defendant wall off every chat as work product. The defendant instead was ordered to show that each chat was made because of the lawsuit and list anything withheld on a log.

The Decision
Reading United States v. Heppner alongside Warner v. Gilbarco, Inc., the court wrote: “Neither decision establishes a categorical rule governing all AI communications.” In other words, using AI does not automatically make a chat discoverable, nor does it automatically make a chat protected. District courts are examining these issues on a case-by-case basis.

The takeaway. AI discovery is spreading past privilege fights, with prompts treated as possible evidence of intent which parties can use to establish their claims and entitlement to damages.